Full Momentum: An HEC-RAS Vodcast (Ep.7). Manning's Roughness for 1D Models

[Music] all right welcome everybody to episode 7 of full momentum and HEC Razvan my name is Ben Carey I'm your host and here with me joining me today is Chris Goodell Chris welcome to episode 7 hey Ben how you doing good good beautiful weather out there today I'm getting to spend actually the last couple days down in Corvallis where I went to college but I'm doing a little project down here I in addition to working so it's good little change of scenery yeah what's keeping you busy like you said just the nice weather we're trying to get outside a little bit and enjoy that as much as we can but our house doesn't have air conditioning so we're doing the thing where we open up all the windows at night and then close them once that starts getting warm in the morning and then just kind of suffer through the afternoon hours a little bit but I know we in Portland are kind of soft when it comes to hot weather and all that I know a lot of folks out there on the East Coast Midwest south have it a lot worse than us so we're a little bit stopped when it comes to that yeah that's true but we can deal with rain though so how about that it's true when it comes to water which we're water guys so that makes it cool well yeah like I said welcome to episode 7 everybody we had a couple really cool topics we're going to touch on today including Manning's values in 1d models as well as a cool tip and trick so we're going to touch on at the end of the podcast but I wanted to start off with a news story and again this is for those of you guys have that has seen our our show we like to touch on relevant topics and water resources or hydraulic engineering world and there recently about a month ago was a really really relevant event that happened and that was a couple of dam breaches that occurred in Central Michigan and so we wanted to touch on those and just kind of chat about them a little bit and give us some good context as to maybe how those would be modeled in heck rad and made some of the tools that we would utilize to potentially have prevented something like this from happening so for those of you guys who are not familiar let me go ahead and share my screen with you the two dams that breached and at the end of May of this year were the envel and the Sanford dams in Central Michigan the envel dam is the upstream dam and you can see the location of that the same for dam is a little bit downstream of that that's located here and then the those two Dame's that breached caused a lot of damage to the town of Midland which is just downstream of the Sanford site so this was a really big deal two dams failing in series it's not something that we see every day it caused a lot of damage so we just kind of wanted to get in to get into this it's okay and it's a unique case study so we're going to show some pictures and some videos and then add let's have some dialog so there was a very serious rainstorm event that occurred upstream of that anvil Dam and this is what the anvil dam looks like currently so you can see it's actually a it's kind of a long and Bank Minh dam with a couple different spillways you have a spillway up here as well as the primary spillway which is which is located here and what happened was is the embankment overtopped because of the the flooding that occurred and if we look at this kind of more of a plain view you can see that Wixom Lake is upstream of the reservoir and the river system feeds into that Lake and the embankment actually overtopped right here and so some of the videos and photos that we're gonna see are gonna show that this really got demoed pretty good and this whole sinuous channel over here actually got bypassed and just this became the primary primary river channel right here so if we compare this to for instance the photo that we have here this gives some good context as to what that looked like afterwards so again we're looking at that same embankment dam here the ever used to kind of meander its way to the left here but because that over bank area was over topped and eroded this became the primary channel so yeah and look at all the look at the bed forms in there or not the bed forms but the the rolling waves standing waves in the middle of that and and you know what that means when you see stuff like that first of all there's a lot of sediment transport happening right there but also you know they look to be to me standing waves which would imply you've got a lot of mix flow going on in there subcritical supercritical happening throughout so yeah pretty dynamic event just by looking at a photograph you can tell yeah and Nike said like you pointed out Krista whenever you have a dam breach one of the really damaging things that can occur is the amount of sediment load that's going to be taking downstream and that's actually what part of the reason why the Samford dam which is downstream of this dam breached is because a lot of the reservoir got filled in with all this sediment so as more and more flow came downstream from the breach had less and less volume to be able to occupy in that reservoir which is part of the reason why that dam was overtopped as well so I think your your identification of heavy sediment load and this is correct I will show you a video to of this scenario or this particular breach it's pretty interesting so I'm going to expand this out again this is a see if I can zoom and I don't think I can zoom it anymore this is that same channel so we have our our primary spillway here the channel would meander to the right and then come back to the the primary channel in here but it said the whole thing got bypassed and this just think really just demoed out really really serious Phil seriously this is a video you guys can see so this is a couple days after so the peak flows have subsided but you can still see there's a good amount of water moving through this is a little bit upstream of the reservoirs you can see how the whole reservoir is drained down that's just a lot of sediment dead trees floating around in that and if we go further in we see a bunny who's boat ramps that are just kind of hanging out to dry because of what occurred there so that's pretty cool gives us a little bit more context as to what happened upstream and then this is this is a really cool video that Chris noted this is again right at so if we open up that photo where the embankment dam breached this is at the start of that that breach occurring this kind of shows that actually current you can see how quickly it occurs and how it looks so I'm gonna go ahead and play the video you guys can see there's some flow coming through initially and then the whole thing just slumps yeah very quickly too right I mean this is what you would call classic mass wasting failure just a big big mass of Earth just slides out of there but notice the flow doesn't start right away and that's because I I still think that there's part of the dam on the inside is still there pointing flow back but then in you know in just a little bit you know the video continued on you would see very quickly the flow pick up at which point I think that's when the breach officially happens and and from there it's just there's no stopping it it's just gonna keep draining until there's no more volume left behind the dam yeah I wanted to show that again just one more time so you can see some of the seepage that's already started occurring you know on the downstream side I'm sure when state dam safety officials saw this going on they knew that there was gonna be some some bad news you can see even downstream just before any of the breach occurred how inundated the downstream channel is so serious rainfall event that led to this again the sloughing that occurs pretty interesting this gives this idea of the failure time we always talk about how long dams take to fail and and that's one of the primary important components to a dam breach models figuring out what that that failure time is going to be and in this instance we're talking you know less than 20 seconds that that occurred so yeah yeah and there it goes now you can see it picking up and I root I kind of wish that the whoever was taking the video kept taking it because for folks like us we want to know you know the widening rates how quickly does that breach grow and what extent and what's happening to the water as that breach is growing so things like that are very interesting for us who model dam breach scenarios to see in you know in video in real life yeah I'm gonna guess that the gentleman who was filming that was was probably overwhelmed with a lot of other things going on and wasn't thinking about how can I make this video helpful to dam breach I know they only would you know stop being selfish and think about us every once in a while you know the the hydraulic engineers yeah totally just to give folks a background to a little bit more on on this event because part of the really interesting part of this is that is it came out shortly after the breach had occurred that the Edenvale dam which was the upstream dam that failed had failed some safety tests or some safety criteria earlier either a year or two before and so there was a lot of controversy around that and ever since then there's been a lot of back and forth as to why the reasoning that this dam failed was the state state Dam Safety officials are claiming that the owner were owners were negligent and that they're negligent behavior led to the failure of the dam the owners are claiming that the state forced them to refill the reservoir to protect some of the reservoir bio fish and mussel species that were in the reservoir so there's this back and forth dispute it'll be interesting to see what the I'm sure there'll be a forensic study done and and a determination will be made I wouldn't be surprised if a SDS Oh was involved with that they were the ones who covered that dam breach in Nebraska that we covered a couple episodes ago so it's always a good idea or you always get the best information from it I'm biased you know source of review so it'll be really interesting to see what comes out of that but does anything else Chris you know in reviewing these videos and photos does anything else strike you as to you know how anything related to Raz relate to this breach well it's an interesting shaped reservoir wouldn't you say not your classical it looks like it's got to tribute two main tributaries coming in yeah and the damn you know almost it's almost like the dam spans two separate independent rivers there thank goodness we're closer and it does and so it's it's a strange setup here one you don't see very often usually you want to put a dam in a fairly narrow Valley area or Canyon and not spanning multiple rivers but just one but here you've got one that spans two so I'm not quite sure why they did it that way and then as far as the makeup of the dam itself I mean you've got it looks primarily earthen embankment but I'm sure that there's some concrete sections there's the spillway support here and here yeah is this say do you know if this is a power producing dam I do not okay yeah I would assume so it looks like there's a little powerhouse here or something that might be connected so I would guess it would be yeah but one of the other things that struck me too is looking at that video you showed if I had to guess I would say that that Slough was maybe on the order of 20 to 30 feet wide but then you look at this final view of the breach that you were showing you suspect may go right there you know that's maybe a couple hundred feet wide wow it's amazing how how much erosion you can get from a breach event like this and I don't know how quickly it got to that point but I imagine it was pretty pretty fast yeah I notice the same thing Chris like in that original video it looked like that first Slough was maybe 20 or 30 feet wide and you can see from Google Earth there that that final breach opening was probably around six or seven hundred feet so crazy crazy but that also gives you a context to when you have a dam that has a reservoir behind it as opposed to like a run a river dam there's so much storage behind here you can start with a really small breach and because of how much water there is to drain it can that can end up really really propagating to something serious yeah and also the other thing I thought about looking at that reservoir the unique shape of it I would almost certainly set this up as a 2d because just look at how wide it is and all that water in the the left arm you know for lack of a better name I don't know what the name of that actual river is but the one on the west side that all has to move laterally parallel to the embankment dam to get to that breach opening so instead of water moving in what you might call it downstream direction in this reservoir it's moving though significantly lateral to that you can get over to that breach opening so that would be a tricky one to try and model with cross-sections I would probably just go right to a 2d area yeah the only way I guess if you really wanted it to be 1d maybe you could do a 1d model here and then another one here and have it connected with the lateral structure across this that'd be my only idea that's actually that's not a not a bad idea and in fact before we had to deem awning and a crab's it's probably what you would end up doing you know you'd have to separate separate independent tributaries and then each one would have its own damn although they would technically be the same dam but you would see out it up and and I think you're right right at that little narrow spot is where you would split it put a lateral structure there and yeah that's a great idea now that you mention that you know maybe that's not a bad way to go here but still you're not going to get that lateral movement of water when you when you model it with 1d reaches but maybe that's not important in this case yeah you know part of it would depend maybe if you're doing just a short short study around a Danville dam a 2d air a 2d model would be a good good approach but you know if you're doing a 1d model that includes envel dam and stanford dam down here and you were doing this whole area in 2d yeah that might take a long long time to run so maybe you're gonna be committed by my computations and maybe you'd have to do something like a 1d 2d model at that point where you'd have to do around your breeches and then 1d elsewhere but that's the medium headcrabs now is there's a lot of flexibility with with how you would approach this particular issue yeah the last thing I wanted to touch on I just thought it was an important point around this breach event there was obviously a lot of news coverage and one of the things that was mentioned is that the was the you know this was the 500-year event the 500 year flood and I wanted to clarify something that was really important related to that and that is the actual hydrologic flood event wasn't the 500 year event it was something less than that but when the dam breached that the breaching of that dam and then the subsequent downstream dam that led to the 500 year flow in the river through the town of Midland and so the people of Midland experienced a 500 year flow that even though it wasn't tied to the 500 year hydrologic event and and those things are definitely important to separate yeah and that really kind of speaks to the difference between say an annual exceedence probability event like the 100 year or the 500 year as you mentioned versus when we talk about dam failures you talk about conditional exceedance probability the condition being that the dam fails and that's a completely different exceedance probability in that case you start with the assumption that the dam is going to fail and then you figure out via a probabilistic analysis what is the probability of getting to a certain level of flow and so you could do that and really you should do that and when you're doing a dam breach study is have a a sound understanding of the uncertainties and trying to do this kind of study and and capture that with a probabilistic analysis essentially yeah go ahead I was gonna say especially when you're talking about dams in series right because the flow going towards Sanford dam is so dependent on what the breach at Eden Vil ends up being and there's so much uncertainty with that you're gonna have a really hard time of nailing down a particular flow to model a Sanford breach at and so if you're do that probabilistically you're able to say all right what's the probability that my dam breach at Hadden bills this size and what would that mean for the probability that my sanford dam failed right and that's when you can really start to unlock some of the power that is incorporated with some of the tools that we've been using I think you were just going to touch on that Chris yeah and don't you think that the owners and the state would have liked to have an idea of the risk of these dams now it's you know hindsight is 20/20 right but you know back when they were talking about hey are these dams under designed are they meeting criteria established criteria for the state if not what is the risk what is the risk that if n bill fails that it will overtop Stanford dam is it Sam sugar Stanford Stanford Stanford yeah Sam for dam and and then what's the the probability of different levels of flow rates downstream in Midland and what does that mean for not only ultimate inundation in Midland but how fast that flood wave will get there and how quickly people have to evacuate and what are some of the more hazardous areas that need to be avoided during an evacuation so all those things can be better quantified doing a probabilistic approach I think and yeah we've been using our software make breach to do this quite a bit and it's a it's a very valuable powerful tool for communicating flood risk and boy if they if they had this ahead of time they may have made some different decisions I wonder yeah especially we talked about again embankment dams because you know there there was a mapped dam breach they had a sunny day and a pmf dam breach that was map through the town of Midland but I think when people see dam breach they think of Oh worst case scenario this isn't very likely to happen yada yada yada but that's a deterministic approach of mapping versus if they had a probabilistic approach maybe they would be able to map not only like the 1% dam breach event but something that would be much much more common and maybe that even the more common event would still have overtopped Sandford and thus caused a bunch of damage downstream of it and maybe that would have you know really shown some red light you know flashing red lights as far as hey even even in a conservative or a less conservative dam breach we're still going to have really big issues here and maybe that would have helped in this particular situation yeah I think it boils down to communication it's all about communicating the reality of the risk that these dams pose to people around them and by providing deterministic maps I think you are a little bit disingenuous and some people may may feel safer than they should and some people may feel a lot less safe than they should and that's what the part of the beauty the probabilistic approach is is it it gives you context to that it tells you if one of my odds of being flooded and everyone understands that they understand probabilities they understand if somebody tells them there's a 50% chance they're gonna be flooded at this dam fails then they know what that means and they know how to prepare for that yeah yeah that's good so if anybody hasn't heard of Mick breach go to our website Kleinschmidt com there's a lot of really good resources there when it comes to the background I'm ik breech how to use it how to download it it's it's a really powerful tool that we would love to see become more mainstream when it comes to dam safety work hey and guess what Ben it's absolutely free so go get it there's no reason not to it's free to download it's free to install the even comes with its own user's manual so I'll give it a try and let us know what you think yeah and if you have any questions about it feel free to leave those in the comment box or send us an email we'd love to to answer those for you and hopefully be part of the solution to help make sure that this type of thing doesn't happen in other parts of our country so yeah awesome well before we get into the main topic for today which is going to be Manning's roughness values and how to use Manning's in a 1d model I'd like to just give a shout out to our firm Kleinschmidt associates for sponsoring this episode and allowing us to to give you guys this valuable information we're really thankful to be sponsored by our firm Kleinschmidt associates who is known throughout the industry as a firm that provides practical solutions to complex problems affecting energy water in the environment you can learn more at Kleinschmidt group com and you can learn more including links to this these vodcasts episodes that we're putting out you can learn more about mcbreezy can more about an upcoming 1d 2d class that we're going to be teaching this fall so go to Kleinschmidt group comm there's a lot of great information not only on H and H stuff that Chris and I are putting out but about the project types and services that Kleinschmidt offers so go check that out yeah go class Smith all right let's talk about n values let's do it so Manning's n values are a critical aspect of any pet grass model it's a really important when it comes to calibration defining the model it can really affect your results especially in a riverine system so we're gonna dive into that today but before we got into the nuts and bolts of how headcrabs is actually employed manning's is implemented within headcrabs I wanted Chris just to give a brief background on what Manning's roughness is and why it's important for hydraulic modeling yeah well it's it's tough to squeeze this into a very relative so again another plug to take our class because we can dive much deeper but if I were to sum up the what the N value is in hydraulic modeling I would say it's an ignorant factor or an ignorance coefficient and what I mean by that is everything we're ignorant of as far as the physics of the movement of water through a river we can lump that into the N value and we can use that as a means to capture the energy loss due to a variety of things but classically the N value is there to describe the energy loss due to the friction of the bed surface so let's say we've got a river and it's got a gravel bed surface that's going to have a certain amount of friction or resistance to flow if we've got another river that has a lot of vegetation in it a lot of trees that have fallen into it a lot of sticks and other debris that's caught up on some of these trees that also is resistance to flow now these are two totally different features in a river but they both provide resistance to flow and so we can wrap those into an N value but there are other things that get wrapped into the N value especially when we're talking about 1 dimensional modeling things that we define in our one-dimensional equations like turbulence like flow separation form losses now in 1d hec-ras modeling if you're doing unsteady you can't you do get losses due to form changes in form in the momentum equation but you know you may have some extreme cases where you have very sudden contraction and expansions that aren't being captured well enough in which case maybe you bump your n values up but yeah typically you're gonna look at the type of stream you have what it's made of how much debris there is how big it is what's the water level all these kinds of things and those will factor into the N value it's a very very subjective exercise which makes me very very uncomfortable for a hydraulic modeling because I don't like subjectivity I prefer that we had a perfect equation to define in values but it's just impossible there's so many things that go into it and things that we just have no way of describing mathematically on their effect on the flow of water through a system yes you could have two really really sharp smart experienced hydraulic engineers that look at the same channel and they come up with a manís value that's different that it's gonna produce different results and so like you said as engineers that can make us uncomfortable we don't like gray area as we like black and white mm-hmm and Manning's before they just just isn't one of those things but at the same time I you know individually I kind of enjoy that aspect of it because it's not just cookie cutter it takes some interpretation and some experience and some conversation with other engineers you then your firm or within your project team to come up with what's best for this particular solution and it also provides some opportunity that we'll talk about here in a little bit when it comes to calibration because there is some opportunity to tinker with that within a kind of a range of reasonability to help calibrate your model we'll talk about that in a second you know you send something very key their interpretation I think that's a very important word in a very good way of describing is this is your interpretation of the roughness in your channel and how that translates to an N value and the only way to really get good at that is with experience it's a very subjective exercise as I mentioned before and the way you can get good at is is by doing more and more of these models and calibrating them as you do them because without calibrating you don't know how close you are to the real end value you don't know if you were right or not so it just comes with experience but there are a number of resources you can use to help you get to a fairly good estimate of an end value and hopefully get really close to a calibrated end value yeah and very good Chris and we'll definitely touch on those on those different resources and references here shortly so good good good little background on meanings let's start with how manning's is actually put into a one DHEC crass model so here I have actually a 1d 2d model we're gonna focus on the 1d reach component of this model here though so normally mani's values is entered into the Edit cross-section editor here so if I open this this up this cross-section data editor you'll see that Manning's input into the dialog box is here for your leftover Bank your channel and your right over Bank your leftover Bank and right over Bank are defined as what is left and right of your Bank stations that are defined so this is what a normal Manning's value setup is in a 1d cross-section so you could either edit the main use values of each individual cross section one at a time in this in this dialog box or if you wanted to edit multiple cross sections at the same time or perhaps redefine Manny's values for an entire reach or section of a reach you can come into tables in the geometric data editor and click on the Mannings and/or K values and this will show you all of your main use values for your whole reach so you'll notice that for every cross-section in this particular reach we have just three defined main use values left over Bank Channel and right over Bank if you if you have more dynamic nests when it comes to your maining values for your 1d cross-sections perhaps you have really long cross-sections that span areas where there's some short grass floodplain and then some heavy forest and then maybe some open land or something like that you want to have some variation within your 1d cross-section aside from just the variation that's provided in the left over bank channel and right over bank you do have the option to horizontally vary your n value not just based on those three different criteria and you can do that by clicking on options in your cross-section data editor click on horizontal variation and end values and then this will automatically change your end values here from your standard left-right and channel values to spatially vary which means that you define the Mannings end value based on its station so you can see that in this cross-section here it starts at kind of a higher station but the end value is 0.07 so this will be the value for your cross-section will be zero point zero seven until you define a new Manning's n value so if we scroll down here it looks like it's zero point zero two eight so that's where the transition occurs there but let's say you have a transition that happens sooner in there you want to break this apart even more you could have a break where this maybe this changes to 0.09 okay if you apply that now the sudden you have another discard ization of mayonese value across this cross-section so you could imagine that if you wanted to you could spend a lot of time discretizing the Mannings value across your cross-sections and yeah we encourage that if your cross-sections are really really large and encompass some of that sometimes it's it's easier more simple just to keep it left-right and channel but there are times when it's important to do this yeah I always like to tell people if you can keep it simple left channel and right overbank that's gonna make it's so much easier to deal with your cross-sections later on and then if you start interpolating if you have this horizontal variation like you see here this is not so bad with only four n values but if you had like seven eight nine n values across this channel and you interpolated with another section that has seven eight nine different n values now you can have as me as twice that number of n values in one in two bladed channel and just get it so messy and weird and so that's why I tell people if you can try to keep it simple here don't go overboard with your definitions but if you have to do it you have to do it I mean if you have farmland over here on you know just on the the floodplain side of the the left bank and then on the other half of that floodplain it's all forest then you know of course those are gonna be two different end values yeah and what Chris was talking about is is in this situation where you have you know six different end values that span your cross-section if I go and actually interpolate cross sections in this one do you reach I'm gonna go to tools' cross section interpolation within a reach I'll go ahead and let's do let's do 30-foot cross sections here I'm going to interpolate between these guys it's kind of taking me a second to do that and now I have a bunch of cross sections in here if we come back to my cross section editor and we look at what some of these cross sections are downstream you can see that they're still really really broken up when it comes to the different cross section or Manning's M values across this so it can get really messy really quick that's why it's always best to start like with me Chris will always say in modeling start simple add complexity where needed Manning's n value is no exception to that rule the last thing I want to talk about for hours pass it back to Chris let me go ahead and delete these interpolated cross sections so we can see a little bit easier is one of the other cool things is if you do have a spatially varied or I should say horizontally varied manes values you do have the ability to adjust those not only by adjusting them in the cross section data editor in this table here but you can do that also in the cross section graphical cross section editor so you can get there by clicking on the graphical cross section editor and that in the upper left hand corner of the window and then you'll see the main these values that are up here that are displayed for your cross section now normally if you just have your standard left over Bank Channel right over Bank you can't edit where these are at but because we've transitioned to a horizontally varied roughness you're actually able to hover over these transitions here and you can see the lines that show up and you can actually drag the and change where these are at so if you have multiple of these this is kind of a quick way to go ahead and just adjust those and this is really helpful if you have interpolated cross-sections and maybe those maining values aren't showing up exactly where you want them to this is a really quick and easy way to do that adjustment so I just wanted to pointed that out too that's another kind of powerful tool when it comes to varying Manning's values across a 1d cross section horizontally now I'm gonna pass it back to Chris because there's another option for varying Manning's values and that is vertical variation so Chris when you talk a little bit about varying Manning's values vertically within a 1d cross section yeah thanks Ben so as Ben was talking about horizontal variation Raz allows you to bury your n values vertically as well and there's a couple different ways that you can do that in headcrabs there's the hard way and the easy way the hard way is built right into the individual cross section editor so if I open that up go ahead your screen - Chris oh yeah thank you you will see let me open that back up so go to the cross section editor you're gonna see under options vertical variation and n values at the very bottom of the list here and when you do that you see for every cross section you can put in a definition of n values for different elevations okay so you might start in this case we'll start at an elevation of 60 put in 60 here and then define all the N values going across the section okay and then maybe you do another one at 65 and then you have 65 here and you fill out the table again now again this is this is a reason to keep it simple because if you just do the left channel or the left overbank channel and right overbank definition of n values and you only have to put three in for each elevation if you've got multiple variation going across well then you've got to do the same thing all the way across here for every single row okay and it lets you go out up two it looks like twenty stations so you got a lot of flexibility to work in here but consider the fact that you have to do this for every single cross-section if you want to use this method and you can vary it based on water surface elevation here or based on flow but let me show you an easier way to incorporate changes in n value vertically and by the way we haven't really talked about this but n values do change vertically yeah the deeper the water is the lower your n value is going to be as you start approaching the bottom of your your section as the water starts to drain out and the depths get lower and lower your n value will actually start to skyrocket it will get very high and at very very shallow depth so you can see n values up in the point two point three range or even higher um so keep that in mind to that point Chris I think it's important to note too that the virk the the reason why you'd want to potentially use vertical variation amazing values if you is if you were modeling a really drastic range in flows like if you're modeling stuff that starts really close to dry and then ends up being you know a flood flow level that would be a really good reason to use vertical variation but a lot of our projects you know we're either only looking at low flow models we're only looking at high flow models in that case there's not necessarily a need to incorporate vertical variation in that case it's just really important to ask yourself you know critically like what type of modeling am i doing and how will that affect my manies value that I choose based on my observed you know channel classification so there is a distinguishment event you're not always gonna want to use vertical variation it's really gonna be wouldn't you're using some serious variation in flow yeah really the two times I ever use vertical variation is is if I'm trying to calibrate to a range of flows maybe I have a gauge somewhere in my river and I want to try and hit all those points both low flow and high flow sometimes the only way to do that is to have vertical variation and n values because an N value of 0.03 at flood stage turns out it might be 0.055 at low flows or low stages and there's no way you'd be able to calibrate the full range of flows unless you do change your own values vertically one of the other reasons is just for stability purposes a lot of times you will start off a model at low flow danbridge model is a great example of this maybe you're doing a sunny day dam breach event and you've got relatively low flows in the channel downstream of the dam and then all of a sudden you drop this dynamic flood event on it and that that quick change from low flow - really - an extreme acceleration you get there locally is really tough to get through if you if you have a low end value and so by providing a higher end value at those lower stages right when that flood wave hits that cross section then you can add some more stability to your model so sometimes I'll do it that way as well so let me show you the other method for a vertical variation and to get there you actually have to go into your plan window your unsteady flow simulation window and this is only something for unsteady flow of course not for steady flow but under options you'll see flow roughness factors okay when you click on that you'll see a table here and the advantage of this over the other method is that you can apply this over a range of cross-sections so you don't have to set this up for every individual cross section so you can pick a range of cross sections maybe I want to do the entire reach here so we'll say all right let's do this entire reach and I'm gonna add it okay so there you can see I've got my Tisa lower reach the entire reach is is selected here and then I just fill out this table and so these n values are going to change based on flow so I'll start at zero and then maybe I go up to you know 500 cubic meters per second or whatever you know if you're using US units cubic feet per second then maybe the next one I do is a thousand oops a thousand then 2,000 maybe 5,000 10,000 okay and then I'll start remember low flows we're gonna have higher end values now these factors are not in values here these are actually numbers multiplied that you multiply the N value by so a roughness factor of 1 we'll just use the N value you have in that cross-section okay and so maybe we'll say you know what I think the N values represent a flow of 5,000 so I'm gonna make this one here but as I start reducing flow my roughness values will go up so maybe it's one point zero five here one point one here one point one two maybe one point one five all right and then something higher may be the end value continues to decrease maybe not it tends to level off but let's just for the sake of argument we'll make this zero point nine five yeah all right so now I've set up roughness factors over the full range of flows and make sure that this last number in here is higher than any flow you're gonna get otherwise razz will have to extrapolate and you never want to have razz extrapolate for you if you can avoid that so that's the roughness factors now if you're in between it's going to use it's going to interpolate in between I believe don't quote me on that I'd have to look that up but I think it interpolates so if like I have 250 it'll just take an interpolated value between these two yeah the disadvantage of that that I see Chris is a lot of times especially if you have a really deep channel and then really wide overbank areas maybe you would want your channel to adjust Manning's wise to different flow amounts but you want your overbank areas to take consistent if you were to use this approach it would change your overbank and your channel areas at the same rate versus if you use this spatially or I should say the vertically varied in the cross section editor itself you could you could keep the overbank areas the same and then you could adjust the channel so I guess it kind of depends on the setup of your 1d cross sections yeah that's a good point that's a good point Ben and the other thing too is besides these flow or ethnic factors I want to point out really quick that you can change n values by season as so maybe in the wintertime where all of the the bushes and then the trees around the river have lost their leaves they're gonna have a little less resistance to flow because you don't have all of that vegetation there and so you might want to adjust your n values down and so this again if you're trying to calibrate over a long term simulation maybe a year long event then you may want to use seasonal roughness factors to account for that I mean it also in reality the viscosity of the water changes with temperature too and so as the water temperature decreases in the winter time then the viscosity goes up and it actually has more resistance to flow so the N value is more pronounced in that respect and so your n values may go up in the winter time as well so it's a little bit tricky you have to think about vegetation and temperature and stuff like that are those measurable though I mean I understand that the viscosity might increase but does that actually have a measurable impact as far as when putting it into your model on mani's you know I would probably wouldn't spend a lot of time trying to get it to that level of precision I'd rather get it close and then calibrate and so in calibration if I see look I've got an N value of 0.03 five and it fits really well in the summertime but I'm just a little bit off in the winter time well then I'll adjust my n value within a reasonable amount to account for that seasonal difference okay but I'm not gonna go through and and compute different n values for different viscosities of water I'm not going to that level of effort - got it it's it's an interesting thing to do and and in fact you can figure out the effect of viscosity on an N value there are equations that do have that are functions of viscosity that you can can use but I yeah I generally wouldn't do that in a practical exercise ok specially if you have a budget yeah and then the last thing I wanted to point out here is this automatic roughness calibration now I have not used this very much I much prefer to do calibration myself I like to have the hands-on approach I like to be able to dial things myself rather than just being hands off and letting rats figure it out some reason I'm not comfortable with that but if you want to try it out there is this automated roughness calibration and so you you would provide in high watermarks and gauges and Raz will actually adjust n values with some parameters that you provide to it like minimum and maximum n values and ranges of cross-sections where you want to apply it and so forth it will adjust those n values to try and calibrate automatically but again I much prefer to do this manually myself so yeah but that's good those are those are three really cool tools Cris that I have very little experience using so I'm glad you were able to show show folks those different options again just more flexibility based on your particular project type the simulation that you're that you're gonna be running the time of year all that can be a lot of flexibility but I'll get built into heck rads that people don't necessarily know right at the top of their heads so yeah I want to point out to Ben before we conclude for the day that there are resources out there for helping you so if you're new to hydraulic modeling or even if you're not new and you're just not quite sure what n value should I use here there are a lot of resources out there to help point you in the right direction I use them on every single project I work on and in fact it's a good idea that you do that because you never want to rely solely on your own experience you always want to have some backup because somebody may challenge you on that in value and you saying ah that's just the N value I think it should be is not good enough you need to have some backup and these this backup will give you ranges and then you want to do sensitivity analysis too and we talked about probabilistic modeling before while n values are very uncertain as we mentioned already there's a lot of subjectivity that's a great way to do probabilistic modeling maybe you want to do some Monte Carlo runs on n values and see what flood risk is when you do uncertainty on his factors but let me pull up a document that I really like to refer to quite often yeah and while you're pulling that up I'll just mention a few resources or references that I use all the time when it comes to Maine's roughness so obviously heck grass has its own reference in the hydraulic manual Relic reference manual it has some recommendations for Manning's values I also reference cha from 1959 which is a solid document that contains some good information as far as what to use for Manning's four different stream types and then the last I'll let you show this first Cris and then I'll go ahead and show the USGS roughness characteristics of national natural channels which I really like to use so what do you have here Chris so this is the the Corps of Engineers engineering manual 1 1 1 0 - 2 - 1601 I've used this since day 1 of my career and this is called hydraulic design of flood control channels and there's a great chapter in here on end value so you go to chapter 5 zoom up to that it's page 47 I believe and here it is chapter 5 methods for predicting end values for the Manning equation and there's a lot of great material in here first of all it talks about analytical methods what if we have grass lined channels what if we've got mobile boundaries it even presents this Cowan method which I I really like because it takes into account all all of the different features that we may throw into this ignorant coefficient in fact the Cowan equation if more than any other equation out there really highlights the fact that we use our n values as an ignorant coefficient because you've got an addition to your n value for surface irregularities for variation in channel cross-sections shape for any obstructions you have for vegetation even a meandering factor so if you have a channel that's a very that's very sinuous has a lot of meanders in it you can adjust it based on this M ratio and so if you read down further on you can get an idea of what the base n values should be and then what all these are different these different additions can be as well if you go down a little bit further you can see how n value relates to the chazy number the Chazy coefficient jz is is commonly used in europe and other parts of the world not so common in the US but I actually like chazy because and I'm sorry this is not the shays ec value by the way that's a different coefficient but let me let me go down to the che Z value which is right here so that's what C is in this equation right here the cool thing about che Z is that notice in the equation let's move up to the simpler form this is the Q Lagaan equation for che Z notice it's a function of hydraulic radius and so we're talking about how the N value really changes with depth and it's up to you to figure out that change well here you can actually calculate it so you've got haz value that's a function of the hydraulic radius this K value is actually a measurement of the the bed material or the roughness elements at the bottom of the bed it's also called the equivalent Sandrock roughness or the Nick erotica value so this is a much more theoretically based equation and so you can get a very good C value here and then convert that to an N value and so if you want to have more theoretical basis to what and I use I highly encourage this cheerleading hooligan equations meant for rigid beds look into that as well there's some other equations that also can figure out a Chazy value for you so that it's a little bit iterative the right Chris because you'd have to compute a Z based on the hydraulic radius but then based on what your shady value is that would change your hydraulic radius so it's a little bit of joining the egg thing right and plus you know in an event to model you're gonna have different hydraulic radii as the water goes up and down and so your C value will change so what you do is you you calculate this for a variety of hydraulic radii and then maybe convert those to n values and then do vertical variation based on that sure so just another good reuse that's cool yeah and this other one that I was looking at before is another one I use quite frequently especially when you're talking about gravel to cobble size material kind of rocky channel bottom streams or in fact this has also been applied for riprap so if you're doing some rip wrap design in the channel this is a great equation it's very simple to use it's just this coefficient right here which you have values for x the nikah rocks of K value the equivalents and roughness and this in strickler equation it can be made equivalent to the d90 of the material or in the case of natural sentiment d50 so just make sure you read how do you apply that Strickler equation but that's a good one it's very easy to use and very quick for for streams that have gravel and cobble type materials or riprap there's one that's not in here that I have to mention because I use it all the time and a lot of that has to do with the fact that Ben and I do a lot of our modeling in mountain streams because we live in an area with a lot of mountains and that's Jarrett's equation so Jarrett's equation is built specifically for mountain streams and I'm gonna Google it really quick because it's really easy to find and my hope I've got my fingers crossed that the first return will be the blog Hey look at that it is so if you go to the res solution you can see all about Jarrett's equation here it is even in the preview of that page you can see what the equation is and it's a function of slope your bed slope and your hydraulic radius so it's very easy to determine and it gives you a great approximation of the N value when you're in a mountain stream and this is much more applicable for relatively shallow flows so when you're looking at low flow events small hydraulic radii this works very well and you would be surprised at how big the in values can be you get n values and mountain streams of you know point one two one five even up as high as point two sometimes so very big n values and that just is a testament to how rough and how turbulent these streams are so if you do have a mountain stream check out Jarrett's equation and use that to see if you're pretty close on your estimates nice that's some good information Chris I'm gonna go the other direction so Chris kinda talked about the analytical approaches for determining some or some additional reference material for Manning's values either the via the corps of engineers manual or Jarrett's equation I really like using visual aids I'm a visual learner that's kind of how I operate oftentimes we have photos of our River reach that we're modeling and it can be kind of hard to look at a photo and then read for instance the referenced and chaough and come up with okay this photo is indicating that my channel is this characteristic I like using photos and so the manual that I use quite often is the USGS roughness characteristics of natural channels I believe this is from the late 60s that this was developed in but they have data for over 50 different channels varying from large open rivers to really really really small steep mountain streams and if you scroll through here you actually see some of these pictures and they have pictures that you can look at me I guess this is exactly what my channel looks like based on the fact that this is what my channel looks like I can read more about how the Manning's values are determined for that type of a channel the different ranges that might be appropriate and again they range from rivers let's go towards the top you know large open rivers like so with small sediment a small grain size in the bottom of the channel to something towards the end of this manual which might be really really steep narrow highly vegetated streams and I really like the visual aspect of this manual so this is another good one again I think Chris would agree with this whenever you're determining Maine's it's always a good idea to you know come up with your own value maybe verify it with some visual aids such as the US yes manual verify it again with some more computational methods like Jarrett's equation or some of the core stuff in the core manual to really give yourself an idea of am I am I close and then if you feel like you're within range with all those different references you can feel really good about defending that choice and mani's value for your model but don't forget to calibrate it and if you can't calibrate it at least do a sensitivity analysis and do what range of results you might expect with a realistic range of n values and a really common range of n values to look at is plus or minus 20% so somebody may say hey I think the N value in this stream is 0.04 but I'm gonna do a sensitivity analysis and I'm gonna also look at plus or minus 20% of point oh four I can't do the math in my head of what that is but so let's just say maybe you look at point O 3 and point O 5 or something if that turns out to be 20 percent I don't know but that's a common way to do it and that gives you a pretty good idea of the range of results you might get as well and and I really like that that book to been in fact I have a hard copy of it on my bookshelf I use it all the time it's a quick way the barns book USGS book the other one though I really like is this roughness characteristics of New Zealand rivers that's put out by NOAA and this is probably my favorite end by a book because it's got great photographs it makes me want to go to New Zealand and check out there are rivers there but it's well well-documented and it tells you all about how they calculated the different n values what conditions they were looking at just like the USGS Barnes book does but this is just another really good one unfortunately I've not been able to find this online so you'd have to buy it I highly recommend it if you've got the budget for it and go ahead and pick this book up it's really good and well worth you adding to your collection of n value options the other thing is just Google I mean get on Google I mean every day almost there's new n value resources out there and you'd be surprised at what you can find just googling and values and rivers and see what comes up yeah good stuff so that we did a great job I think covering you know how many is input into Hector as some of the different ways you can edit both horizontally and vertically Manning's values for 1d models and then some of the reference material that we'd recommend using when coming up with that so before we close out Chris is there anything else that you'd like to to mention well I have to say n values are probably one of the trickiest parts of hydraulic modeling and we're going to talk about how we apply it to 2d modeling in the next podcast this was all about 1d modeling it's even trickier in 2d modeling but just remember to keep it simple that's gonna make it easier have some backup and it's just gonna take experience to be good at selecting n values and until you get there make full use of all the resources out there that we just talked about and anything else you might be able to find just by googling around on the web yeah absolutely cool well that was a great discussion Chris thanks for showing folks some of those kind of new maybe more hidden ways that you can kind of customize a 1d model to incorporate some of the complexity that that mains values have so yeah this is a great topic really enjoy talking about n values and like like I said before there's a lot more than we can fit into this one vodcast so if you want to know more about it yeah think about taking a training class and and why not take one from me and Ben right then absolutely for those you guys who aren't aware yet we're gonna be teaching another class in the fall starting in early October leave October 7 so look at our website Kleinschmidt group comm look at the heck rads blog all the announcements will be there if they're not already there now they'll be up there shortly we'd love to have you join us and really get into the nitty gritty on things again this these vodcast is kind of we're just scraping the surface on how much information there is in those classes we go a lot more in depth so I guess with that being said thank you everybody for joining us this has been full momentum and she's here as vodcast [Music]

Channel: The RAS Solution

Views: 5,932

Keywords: Hydraulic Modeling, HEC-RAS, 1D Modeling, Engineering, Water Resources, learning, modeling, dam breach, dam failure, Manning, n value

Id: 51EEg5N4lx0

Length: 60min 35sec (3635 seconds)

Published: Thu Jun 25 2020